ErrataBench results for GPT 5.6 are out, and 5.6 Sol beats Fable 5 basically on all fronts. Very solid model release from
@OpenAI. 5.6 Sol seems to perform best with Medium reasoning. More in thread ⬇️
ErrataBench results for GPT 5.6 are out, and 5.6 Sol beats Fable 5 basically on all fronts. Very solid model release from
@OpenAI. 5.6 Sol seems to perform best with Medium reasoning. More in thread ⬇️